To date, most cell types in mammalian organs have been identified and categorized based upon histological and histochemical features. It has become clear that within these broad histological categories, there reside subpopulations of cell types that exhibit distinct gene expression profiles and consequently have unique functional attributes. We hypothesize that functionally-distinct cell types comprising the human prostate and bladder can be identified and isolated based upon unique profiles of proteins found on the cell surface or in secretory vesicles. These proteins will comprise what we propose to term the Urological Cluster Differentiation (UCD) system. Through an iterative process, we propose to sequentially isolate increasingly pure populations of genetically-homogenous cell types that comprise tissues representing normally functioning organs. We further hypothesize that cells expressing surface or secretory proteins deviating from cell profiles established as normal, represent additional cell types with pathological behavior (e.g. malignancy, scarring). To test these hypotheses we will: Aim 1. Characterize the repertoire of cell surface and secreted proteins expressed by rt/sto/og/ca//y-distinct cell types comprising the epithelium and stromal constituents of the human and mouse prostate and bladder. Aim 2. Develop and validate a panel of Urological Clusters of Differentiation (UCD) Markers suitable for identifying and isolating distinct cell types comprising the prostate and bladder. This will comprise the first iteration of the Urological Cluster Designation (UCD) system. Further iterations of UCDs will be developed based upon the heterogeneous expression of additional UCDs within the original 'homogeneous' populations. Aim 3. Characterize functional differences in cell types defined by the Urological Cellular Differentiation (UCD) System. These studies will involve the evaluation of in vitro growth characteristics, cellular gene expression, organ topography, inter-species conservation, and transgenic animal models.